Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

   Home   Help Search  
Pages: Prev 1 [2] 3 Next   Go Down
  Print  
Author Topic: HF mobile ant position  (Read 7345 times)
W9PMZ
Member

Posts: 573


WWW

Ignore
« Reply #15 on: December 14, 2010, 12:58:52 PM »

"because you have no way of accurately measuring the ground loss."

Just because you may not have a meter that when connected to two terminals results in the ground loss value does not mean you cannot infer this value by making other related measurements.

Engineers and sciencetists do this all the time...

73,

Carl - W9PMZ
Logged
K0BG
Member

Posts: 9860


WWW

Ignore
« Reply #16 on: December 14, 2010, 01:02:21 PM »

I agree Carl, but the term accurate is what's in question. As I stated, we're trying to assume it is a linear function, when it is anything but linear.
Logged

WX7G
Member

Posts: 6033




Ignore
« Reply #17 on: December 14, 2010, 03:09:26 PM »

Bottom line: The ground loss resistance does not need to be known to measure short antenna radiation efficiency. Only the total of all losses need be known. This is the measured input resistance.  

Loss in the ground near a short antenna manifests itself as an increase in the antenna input resistance. When we quote a number for this it is input referred, meaning it is equivalent to an actual resistor at the antenna input.

A second method to measure radiation efficiency of a short antenna is the measured VSWR bandwidth. One takes a lossless antenna simulation and adds series loss resistance at the antenna input until the simulated VSWR bandwidth is equal to the measured VSWR bandwidth. For a base-loaded antenna this is entirely accurate. For a center-loaded or top-loaded antenna an error is caused by the loss resistance being placed in one spot rather than two (at the input for GND loss and at the coil for coil loss). But, as pointed out in an earlier example this error can be less than 0.5 dB. Alternatively, one distributes the loss resistance more realistically by placing an estimated loss resistance at the center or top loading inductor.

« Last Edit: December 14, 2010, 03:41:10 PM by DAVE CUTHBERT » Logged
K0BG
Member

Posts: 9860


WWW

Ignore
« Reply #18 on: December 14, 2010, 04:06:04 PM »

This is from my web site:

Quote
In the Technical Correspondence section of the September 2006 issue of QST (page 57), are a few paragraphs written by Dr. Jack Belrose, VE2CV. Jack explains how to use an antenna analyzer and EZNEC to calculate the efficiency of a mobile antenna. The basic premise is to compare the measured input impedance of your mobile antenna, compare it to the modeled impedance given by EZNEC, and then adjusting the coil Q (resistive loss) until the two impedances (measured and calculated) equal. Then reading the programs calculated radiation efficiency.

This gives you a close approximation of the efficiency. This is essentially what you're trying to arrive at, but in all due respect David, you're going about it incorrectly.

Again, and again, you can not use the VSWR bandwidth to calculate efficiency. It just doesn't compute! There is even a formula out on the net (it was once on Phil's, AD5X, web site) which supposedly ended up with the figure. Tom Rauch, W8JI, responded to this very subject in a post a few months ago to an article about this very subject.

From me, not Tom, I suggested you think outside the box. Remember, the reactance doesn't change linearity on both sides of the resonant point. If for no other reason, this should give you some idea of the futility of using bandwidth. Apparently, it hasn't. What's more, the change in reactance to each side of resonance, it directly dependent on the antenna's structure. What you arrive at with once specific case, cannot be applied to another, wholly different design.

My point about the ground losses stands. They are caused by the standing waves between the superstructure of the vehicle in question, and the surface under it. It should be obvious that you can't measure them. Sure, I agree it is what causes the ground loss in the first place. However, there are a few gotchas. One, you can't measure them, and you can't assume them. They are frequency dependent, vehicle dependent, surface dependent, and are even effected by the existing STP.

Alas, there is no definitive way to measure, compare, or even speculate the efficiency of any given antenna, except to measure its actual field strength. But even that must be done in a all-factors-normalized fashion. Few, if any, amateurs have the wherewithal to do so. 

By the way, I'm still waiting for the list of patents you have. So far, I can only find two which relate to a timing circuit.
Logged

KI4SDY
Member

Posts: 1452




Ignore
« Reply #19 on: December 14, 2010, 05:56:19 PM »

I am glad to see that someone is finally "footnoting" where he is copying his information from. That is the ethical thing to do. Wink
Logged
W9PMZ
Member

Posts: 573


WWW

Ignore
« Reply #20 on: December 15, 2010, 02:04:14 AM »

"Again, and again, you can not use the VSWR bandwidth to calculate efficiency"

So would someone please explain the method of the VSWR bandwidth method, and providing the equations?

Regarding ground loss.  Standing waves?  Isn't this just about the capacitance between "ground" and the vehicle?  If so, then the ground loss would be linear and dependent upon frequency.

73,

Carl - W9PMZ
Logged
K0BG
Member

Posts: 9860


WWW

Ignore
« Reply #21 on: December 15, 2010, 04:52:52 AM »

Once again Carl, you're correct, but with a caveat. The coupling is not consistent, and changes with frequency, as we already know. However, the change isn't linear, and in fact it is possible to have a higher ground loss at the top of 40 meters, than you do at the bottom of 40 meters. It is also not uncommon to have more loss on 40, than you do on 80. Seems weird that's for sure.

I forgot. Here is what I say about the formula on my web site:

Quote
By the way, there is a formula circulating the Internet which states that antenna Q is equal to 360 times the frequency in MHz, divided by the 2:1 VSWR bandwidth in kHz. One has to assume they mean antenna system Q, but that's not a given. While this formula might give you a comparison between antenna A and antenna B (all else being equal), the actual Q of the antenna (system or otherwise) requires a textbook-full of formulas, and a lot more information than just the 2:1 bandwidth! Fact is, this formula is no more specific than the number of DX contacts a specific antenna garnered.
[/i]
« Last Edit: December 15, 2010, 04:55:56 AM by Alan Applegate » Logged

W9PMZ
Member

Posts: 573


WWW

Ignore
« Reply #22 on: December 15, 2010, 05:34:08 AM »

"Remember, the reactance doesn't change linearity on both sides of the resonant point. If for no other reason, this should give you some idea of the futility of using bandwidth."

I think that there is a reason, if you look at the power response of the system you essentially see a bandpass response.  Circuits 101 tells you that the center of the response is the geometric mean and not the center of the two points that are used to define the start and stop.

73,

Carl - W9PMZ
Logged
WX7G
Member

Posts: 6033




Ignore
« Reply #23 on: December 15, 2010, 07:09:01 AM »

I think an article written by me will help educate. Whille this subject got it's start at eham.com a better place to publish the article is at eeweb.com.
Logged
K0BG
Member

Posts: 9860


WWW

Ignore
« Reply #24 on: December 15, 2010, 07:36:22 AM »

David, I'm still waiting for you to post your patent numbers.
Logged

W9PMZ
Member

Posts: 573


WWW

Ignore
« Reply #25 on: December 15, 2010, 08:12:59 AM »

 PAT. NO.  Title
1 6,675,313  Method and apparatus for reducing signal timing skew on a printed circuit board 
2 6,526,519  Method and apparatus for reducing signal timing skew on a printed circuit board 

found two...

maybe more but i am not really that familiar with the search tool
Logged
K0BG
Member

Posts: 9860


WWW

Ignore
« Reply #26 on: December 15, 2010, 11:35:21 AM »

I found those two myself, but he earlier said he has 6 or 8 whatever it was, and several pending. He has since edited the post, and removed the reference. Apparently, he doesn't have the number he said he had. Hum....
Logged

W9PMZ
Member

Posts: 573


WWW

Ignore
« Reply #27 on: December 15, 2010, 01:09:32 PM »

maybe not the case, could have multiple names on the patent.  nonetheless...

WX7G makes the statement,

 "We can add loss resistance to make the 3:1 VSWR bandwidth 290 kHz:"

then K0BG makes the statement,

"By the way, there is a formula circulating the Internet which states that antenna Q is equal to 360 times the frequency in MHz, divided by the 2:1 VSWR bandwidth in kHz."

Not understanding fully either method at this time, perhaps apples are being compared to oranges...

73,

Carl - W9PMZ
Logged
K0BG
Member

Posts: 9860


WWW

Ignore
« Reply #28 on: December 15, 2010, 02:56:24 PM »

Don't take it out of context Carl. The formula is bogus, and relates back to some article published on the Antennex site a few years ago.

I looked for it, but didn't spend a lot of time, but Tom, W8JI, commented on the formula a year or so ago. I don't remember who posted the article, but the subject was similar to this one. Tom's answer was very explicit, and if I do find it, I'll add it.
Logged

N3OX
Member

Posts: 8847


WWW

Ignore
« Reply #29 on: December 15, 2010, 05:29:21 PM »

I think we need to differentiate between specific, oversimplified, faulty procedures and more sophisticated comparisions of feed measurements and models using a more general technique.  If I had a way to know the antenna's radiation resistance and reactance referred to the feedpoint at every frequency on the 40m band, and I did a high accuracy measurement of the feed impedance and compared to the KNOWN lossless case, I could get an excellent estimate for the efficiency of the antenna across the band.

This would be true even if the loss resistance was a function of frequency, because I'm not just simply comparing something like the 2:1 VSWR bandwidth... I'm actually predicting the resistance of the antenna in the absence of ohmic losses and measuring the resistance with ohmic losses.  I can calculate antenna efficiency as Rrad/(Rrad+Rloss) with Rrad (the radiation resistance) and Rloss (the loss resistance) referred to the same point in the antenna.  In this discussion, that point is the feedpoint.   Whether or not you can use this fact to estimate antenna efficiency hinges entirely on whether or not you have an accurate prediction of the radiation resistance or not.

Assumptions of how the loss resistance changes with frequency only cause errors if you make those assumptions.  So maybe the loss resistance doubles going from 7MHz to 14MHz.  Fine.   That can be incorporated into the model, and you can see if this is happening by comparing model to measurement.  If I recall correctly, a lot of the "you can't use feed impedance to measure efficiency" discussion centers around some examples with sparse radial systems where feed impedance AND bandwidth AND field strength went UP at the same time as radials were added.  I can imagine a few ways that this would happen (current max isn't at the base of the antenna for some reason, and moves closer as radials are added... ground system reactance requires re-resonating the radiator by changing its height, etc).  This is an important counterexample to the idea that you can ALWAYS associate REDUCED feedpoint impedance and narrowed bandwidth with a more efficient antenna.  You can't ALWAYS do that.

But I don't recall anyone bringing a detailed measurement/modeling comparison into those discussions.  I don't think the disagreement was with a model, I think the disagreement was with a too-simple assumption that more radials ONLY change the loss resistance.  I've already come up with a couple of model examples in EZNEC where I can get the total feed resistance AND the radiating efficiency to go up at the same time when I add radials close to the ground and re-resonate the vertical.  But this effect is associated with changes in the radiation resistance referred to the feedpoint as the antenna is re-configured physically and the current distribution changes.

It's important to constrain models with measurements to make sure you didn't leave something important out of the model.  And I think that it's important to really understand the quantitative limitations of what we can do when parts of the antenna are close to the ground.

But at the same time, I think we're getting a bit lost in the details and not really discussing what's at the heart of WX7G's claim: the radiation resistance of an antenna mounted down low on a vehicle is modified by the vertical component of the currents flowing on the vehicle.  This is a reasonable hypothesis with modeling and general theoretical evidence to back it up.

As far as actually doing the model/measurement comparison based on impedance?  You can imagine several levels of comparison.  The 2:1 SWR bandwidth (a two-point comparison) is simple, but you could also check the shape of the SWR curve in comparison with model.  In EZNEC, it would be very tedious to generate a model SWR curve with a frequency-dependent loss resistance for comparison to measurements, but it could be done.  In other software it would be straightforward.   
Logged

73,
Dan
http://www.n3ox.net

Monkey/silicon cyborg, beeping at rocks since 1995.
Pages: Prev 1 [2] 3 Next   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!